There's a shop near me that has loads of U-type DIN ports, let me know how much you need and I'll send some your way.
Printable View
Utsource is a total gamble. There's nothing more frustrating than soldering in a surface mount chip just to find out it doesn't even work. I'm never going to buy anything from utsource again. I've had nothing but working stuff so far from hong kong sellers.
Please! Let me know how much it would cost for two U-shaped 8-pin DINs.
Depends on what you're looking for, really. There are some things I'd buy from them, but there are other things (YM2413s and YM2413Bs) that I would NEVER purchase from UTSource ever again and I strongly advise anyone who's looking for a YM2413 to do the same (for EVERY seller in Hong Kong/China, really. Word of advice: ALWAYS ask for a chip made in 1991 or earlier - the later ones may be counterfeit, and if they're made later than 1993, they're DEFINITELY counterfeit unless it's a YM2413B, in which case it's not so clear when counterfeit YM2413Bs started being produced, but they were around in 2001, so remember to purchase YM2413Bs made before 2001).
I just tried this on my Neo Geo... it didn't work at all. The video is stable or a second or two, then starts to scroll like it's losing its sync signal. I'm going to take the circuit out and plant it into a virgin VA3 tomorrow to see if there's any difference, but for now, I'm heartbroken.
Sounds like what happened with my faulty BA6592F. It worked for a few minutes before the video went completely out of sync and the blue progressively got weaker. The faulty chip would actually work correctly if I bypassed the amp on the Luminance or if I directly put the sync on the Luminance output, in which case, anything gray or white would show up black (a clear sign there's no luminance being output from the transistor at all). This would, however, make the picture much darker and more saturated, so you would need to put stronger resistors at the red and blue outputs after they're amplified.
I must also ask: where are you taking the RGB from? The video encoder's RGB inputs or outputs? Remember this: you must use clamped RGB for my circuit to work correctly.
RGB out. I'm half tempted to desolder the av out jack and wire it from there. Provided the circuit works on the Genesis, that is.
The circuit works no problem on my VA3 Genesis Model 1 I modified for a second RGB output from the EXT port minus the aforementioned video problems (most likely caused by either a bad cable or a damaged trace on the Genesis' PCB). I'm gonna ask Apolloboy to ship me some U-shaped 8-pin DINs so I can test my RGB to Component box on an unmodified Genesis.
Tip: NEVER bend the pins on the BA6592F. It seems very sensitive to bent pins, and I think this is what caused the blue to fail on the first BA6592F I used in my RGB to Component box and what caused the red to fail on the second BA6592F I desoldered off of a dead Super NES. As soon as I swapped encoder on my RGB to Component box, the sync issues disappeared and the blue was perfect. So long as I don't plug in my EXT-port-turned-RGB-output-modded VA3 Genesis Model 1, the box works perfectly, and let me tell you something else: be sure to glue down the encoder then solder directly to the pins. Don't use a surface-mount to DIP board; I always got crap out of these.
Yeah I glued the ic straight to the board and worked from there, although I may have damaged something in the process... I'm currently on the couch recovering from a possible heat overexposure, but I have some time off so I can get to it soon and see what happened. Good thing about glue is I can pop the ic off and put another in its place easily.
So, thanks to Ace, I finally got this working on my Genesis, with the Neo Geo to follow shortly.
In short, this thing looks great. I was a bit concerned about the color temps on my test TV, but once I hooked it to my RPTV, all my fears subsided - it looks emulator perfect. The same can actually be said for my 40" HDTV - it can apparently take a 240p signal. The best thing about that is, is there is no lag time in button and d-pad responses. You can see the overscan, but that's not a dealbreaker. Although, my HDTV seems to be picky about it's L/R sound channel seperation - the left is really muffled. I'm going to connect my Sega CD and see if that remedies the issue, and if so, a cap replacement is in order. Although, for now, I'm glad it works! :)
+1 to Ace, since I can't rep him any more at the moment.
Can you share with the rest of us what the eventual fix was?
Well, I'll try and make it as soon as I get the chips, and as soon as my Va2 revives.
The sync issue I cleared up by using a 2N3904 transistor instead of a 2SC945, but I suspect there was a bad solder joint somewhere along the sync line.
I also had B-Y and R-Y reversed, so that was easily fixed. I've created a PCB using Pad2Pad using Ace's design; we've talked about a batch order, but we'd need at least 100 for the order to be cost effective.
You're looking at 2.75 per board if we do 100, but Ace has final say in this since it's his circuit.
Oh, yeah, I totally see your point. That cost is just for the board and shipping - depending on where you get the other parts, you're looking at around another 5-6 bucks plus shipping on top of that if you buy in bulk. I had been wondering how to get 100 boards ordered, and a group buy seems like the best way - I just dont want to overstep my bounds, you know?
Hey guys - I know you are currently discussing the circuit (and potential PCB) for the BA6592F-based RGB-YUV project, but I wanted to comment on Ace's post back on page 6 regarding adding component video to a launch model SNES as I have that one.
I pretty much followed what was here - http://www.sega-16.com/forum/showthr...l=1#post483965
My NPN was your typical 2N3904 from Radio Shack. After completing the circuit yesterday, the colors seemed a bit over-driven on my plasma. I ended up adding a 75 ohm resistor to the end of the output for the Pb & Pr, and it scaled the colors back to something much easier on the eyes (at least when paired with my display). I do still have a small issue of the faintest of vertical bars, however my circuit is also very much in prototype mode as I have fairly long wires hanging outside of the SNES.
A few questions -
1: Why put 66 ohms on the Y output? Is the Y output for S-Video (I grabbed my Y signal from the multiout connector) not spec'ed the same for S-video & Component video? I would just assume use the signal as it was, however I stuck with the 33 ohm x 2 series take that Ace mentioned in his post.
2: A user on nesdev mentioned going a different route for the Pb & Pr outputs by using a PNP transistor & going through your typical 220 uf + 75 ohm on your output line. Any thoughts on this? http://nesdev.parodius.com/bbs/viewt...&postorder=asc
3: I suppose this is mostly for Ace, but which outputs did you use on your SNES? It looks like you could panel mount RCA connectors, but not without tearing out some of the RCA shielding between the sound box and the back of the case. I suppose you could have abandoned this in favor of your external solution, but I think I am going to pursue the direct RCA connection route if possible.
Possibly some more questions/comments to make, but I am happy to finally be registered and posting on sega-16. Ace asked if any with a launch SNES have tried this mod, and I am raising my hand.
Long story short, it looks pretty good at the moment.
Seems high, how many sq in is the pcb?
I've laid out numerous pcbs and had them fabbed with Dorkbot, they do a medium run order that's $1/sqin! Great deal for US fabbed, 6/6 spec, double sided purple solder mask and double sided silk screen, and gold plating!
Are you trying to keep it through hole except the ba6592/4?
Let me know if you need help, i'm totally willing!
@Ace
I couldn't get this to work in place of the jrok encoder on a PC10 modded nes, any idea what I should look for?
I tried numerous variations of the circuit, but did not try a different 6592 yet.
I know the nes works through the jrok though, here's a vid.
When I first got Component video out of my launch-model Super NES, I had done many changes to the TV's settings via Service mode without really paying too much attention to my TV's default settings. I managed to get the settings back to what they might have been with the help of someone who has the same TV I have (Electrohome 27E510 CRT), and 66ohms actually makes the picture too dark and a bit too saturated. For some reason, direct Luminance in S-Video looks fine (this is how my S-Video cable is wired), but when using direct Luminance through Component, the colors seem too dull. A single 33ohm resistor is actually the way to go here.
I might try this at some point, though I'm not entirely sure why a PNP transistor is used in place of an NPN transistor. Still, I've got a few 100uF caps laying around, so I'll give this a shot in my external RGB to Component box.
I used the multi-A/V out for Component. Eliminated the red and blue outputs and wired B-Y and R-Y there. I was none too impressed with the Super NES' RGB, so not having RGB out of the multi-A/V out is not a big deal for me. Besides, I've got an RGB-modded Super NES Mini with VASTLY superior RGB to what a launch-model Super NES is capable of.
Soon, though, I will attempt to make a converter based off the Nintendo S-ENC/ROHM BA6594F encoder. I had tried something at first using the S-ENC from my SNS-CPU-GPM-01 Super NES (though now, the video has a MASSIVE yellow tint, so I'm going to remove the encoder in its entirety and just use RGB on this system as from what I've seen in a schematic showing the RGB output of the SNS-CPU-GPM-01 and SNS-CPU-GPM-02 motherboards, the RGB is amplified separately from the encoder, so it should still work even with no encoder in place).
What does the NES do when you run its RGB into my converter circuit?
Most of the trials yield only black screen with no indication that the tv is trying to accept any signal.
Only one trial yielded an image, but it was dark and half scrambled, weird.
I'm gonna look at it again this week and make note of which trial yielded the scrambled image.
can someone make one for me? I have a SNES.
If you have a Super NES with the BA6592F or S-ENC encoders, you could just use the Component outputs of the chips already in the console.
By the way, while sifting through UTSource, I discovered they actually stock the Nintendo S-ENC encoder. I'm gonna desolder the one on my SNS-CPU-GPM-01 Super NES (I'm getting all sorts of video issues since resoldering the chip in place) and see how to get good Component video out of it. I've asked for a quote to see how much one of those chips costs, and if the S-ENC could give the same results as the BA6592F with less/cheaper parts, I may ditch the BA6592F and use the S-ENC instead.
EDIT: Bummer... UTSource doesn't have any S-ENCs in stock. They do, however, have 100 BA6592Fs in stock for $1 each. So, with Lastcallhall's quote for 100 PCBs at $250, this would bring up the cost of parts I can't get locally to $350. The capacitors, transistors and resistors are all easily attainable locally, so I would like to get a group order for those 100 PCBs and 100 encoders.
Im in for $70.00 worth of parts (20 encoders and boards).
Also, shipping on the boards will be $25.00 (we can divide that after we get the order filled), and will take 14 days for manufacture. That should be the same time for the encoders, based on my last experience with UTSource. Not sure what their shipping would be, but I can't imagine it would be much.
Ace, did you want to start a group buy thread?
I'd probably be interested in 2 of them for my Saturn and Genesis if you get boards made for it and it comes out cheaper than getting those SCART to YUV convertors.
Depending on the final price, I could go for 5 to 10... at least 5.
If you're willing to use Wester Union or anything like that, I'll be in for 10 encoders.
I can get the boards locally...
Noted.
Shipping's $4 from UTSource, but sometimes, it can be SLOW AS HELL! I once had to wait for just over 3 weeks to get two YM2413s from them (and wait was nothing more than a waste of time because one chip was fake and the other didn't even work).
I didn't think of that, but I will start one at some point (not right now as I want to test the compatibility of my converter circuit as much as possible).
Speaking of which... Trekkies, you may want to make note of this: the converter works perfectly on the Sega Saturn. I just built a custom dual RGB/S-Video cable using a broken S-Video cable, a leftover DB9 (this carries the RGB) and some leftover red and white RCA cables. I would have never imagined the Saturn looking this good; 3D games looks really nice (they don't actually look too blocky - textures look REALLY good) and the 2D games... wow. The Sega Ages releases in particular look like if you were using an arcade monitor to play the games on; VERY vibrant colors and a very sharp picture. I've got some pictures of After Burner II, Daytona USA Championship Circuit Edition and Sega Rally Championship on my laptop, which I'll show later on as I don't have access to my laptop right now.
I tell you, after having used the Saturn in S-Video for so long and now using the Saturn in Component, S-Video just doesn't stack up to Component. In S-Video, the picture is quite a bit blurrier, dithered graphics tend to flicker, there's less color variation and 3D games look blockier (as a small note, my Saturn doesn't have the usual CXA1645, but rather some Fujitsu encoder whose model number I can't remember).
I never could get behind the svideo fandom. Beats composite obviously, but...
I can definitely see a difference when using Component on my Electrohome 27E510 compared to S-Video. Picture's sharper and the colors have a lot more variation (it's not like all the colors are at maximum saturation like in S-Video or Composite).
Did you use an external power source when you tried it on the Saturn or did you use power from the +5V pin on the AV port?
@Ace: So far, it's worked on my Master System, Genesis, and I'm pretty sure it'll work on my Duo once I get the RGB amped. I still want to try it on my Saturn and DC, along with a couple other consoles.
The Neo Geo, however, is perplexing me. I've taken a working circuit out of one of my genesis units and placed it into the Neo, and I still get a vertical scroll. I've posted a video on Youtube:
I've also posted this information on neo-geo.com, and I'm waiting to see if anyone has an idea. Additionally, I've protoboarded the circuit, and I plan to take it into my school lab so I can work with it on a oscilloscope to see why this sync signal is so screwy. It's strange, because the Y signal off of the CXA1145 works just fine on the s-video jack I installed. I wonder if the sync in is what's throwing things off... An oscillator may be needed for this one.
Got a quotation from UTsource, about the S-ENC.
I quoted for 10 chips, and they gave me a price of $2 each.
WTF??? I quoted for 1 chip and I got no price whatsoever.
Still using an external power source. I wanted universal compatibility with my box, so I just made it work with an external DC 5V power supply.
Good to know it works on the Master System. Is this a system with a CXA1145 or V7040 encoder?
It pretty much looks like Composite Sync is completely missing. That's what happens on my TV when I don't have any Composite Sync; if the picture's dark, it doesn't sync, but if it's bright, it looks like it wants to sync, but it doesn't.